3582 J. Agric. Food Chem.

2003, 51, 3582−3585

The Effects of Frozen Storage Conditions on Lycopene Stability

in Watermelon Tissue
WAYNE W. FISH* AND ANGELA R. DAVIS
United States Department of Agriculture, Agricultural Research Service, South Central Agriculture
Research Laboratory, P.O. Box 159, Lane, Oklahoma 74555

The purpose of this investigation was to evaluate the rate of deterioration of lycopene in watermelon
tissue during frozen storage, because little is known about the stability of watermelon tissue lycopene
under cold storage conditions. Heart tissue from each of nine individual watermelons was stored at
-20 or -80 °C as either small chunks or puree and periodically sampled over a year’s time. Initial
freeze-thaw experiments indicated that a small percentage of lycopene, ∼4-6%, degraded during
an initial freeze-thaw. Analyses of the samples showed a loss of ∼30-40% lycopene over a year’s
storage at -20 °C and a loss of ∼5-10% over the same period at -80 °C. Lycopene was slightly
more stable in pureed compared with diced watermelon tissue at -20 °C, but not at -80 °C. The
kinetic data were best fitted by application of two simultaneous, first-order decay processes. HPLC
analysis of the samples after a year’s storage suggested that β-carotene was more stable during
storage at -20 °C than was lycopene.

KEYWORDS: Carotenoids; lycopene; lycopene stability; watermelon; watermelon frozen storage

INTRODUCTION underestimated as a result of degradation during storage, and

Lycopene, a fat soluble carotenoid, is a precursor of β-caro-
tene (1) and has at least twice the antioxidant capacity of
β-carotene (2). A number of epidemiological studies have
suggested that positive health benefits can be derived from the
consumption of diets high in lycopene (3), although a consensus
for either its beneficial or detrimental role in the modulation of
carcinogenesis remains to be established (reviewed in ref 4).
Like the more widely recognized lycopene source, tomato,
watermelon is a rich source for this carotenoid. Because of its
value as a phytochemical, many breeders want to maximize
lycopene content in their watermelon breeding lines, and growers
want to utilize production methods to increase lycopene content.
The need for specialized, expensive equipment, toxic organic
chemicals, and trained technicians for the current lycopene
detection assays makes quantification of lycopene content
impractical for some breeders, producers, and researchers. Thus,
in many cases, samples have to be collected, stored, and shipped
to a laboratory equipped to perform these analyses.
Tomato lycopene has received virtually all of the investigative
attention with respect to its stability under various processing
conditions (5, 6, 7) and storage conditions (8). Conversely, there
is only limited information published on the stability of lycopene
in cut watermelon tissue stored at refrigerator temperature (9).
It may be that the stability of lycopene will be different in the
two systems, because of differences in the attending tissues and
compounds with which the lycopene is associated.
Two questions that the watermelon industry must address are
as follows: (1) has the lycopene content of watermelons been
* To whom correspondence should be addressed. Phone: 580-889-7395.
Fax: 580-889-5783. E-mail: wfish-usda@lane-ag.org.
(2) what are the optimal storage conditions that will minimize
lycopene deterioration during storage? The purpose of this
investigation was to evaluate the effects of tissue preparation,
temperature, and duration of frozen storage on total lycopene
(all-trans plus cis-isomers) in cut watermelon tissue, with the
goal of finding practical long-term storage conditions optimal
for lycopene preservation between tissue sampling and analysis.
MATERIALS AND METHODS
Sample Preparation. Watermelons used for this study were from
the 2001 crop grown at Lane, OK, under standard cultural practices
that followed Oklahoma State Extension Service guidelines (10). Melons
were cut and tissue was removed the day after harvest. All steps from
the time watermelons were cut lengthwise were performed in subdued
lighting at room temperature, unless stated otherwise. Heart tissues from
nine watermelons (Citrullus lanatus cv. Sangria) were individually
processed. The excised tissue was cut into approximately 3 cm3 chunks
or smaller. The collective chunks from an individual melon were mixed
by hand to attempt to obtain as homogeneous a distribution as possible.
One-half of the chunks were divided into ∼10-g samples in zip-closure
plastic bags for the chunk storage studies. The other half of the chunks
from each melon were placed in a blender and homogenized to a puree.
The puree from each individual melon was proportioned out in ∼10-
mL aliquots in 15-mL screw-top plastic conical centrifuge tubes for
the puree storage studies. We elected not to exclude oxygen from the
frozen samples for two reasons. First, nothing is known about the
susceptibility to degradation of lycopene in watermelon tissue at low
temperature under atmospheric O2, and lycopene susceptibility during
processing or storage appears to vary with the attending tissue matrixes
(5, 8). Second, sample processing in preparation for storage and its
influence on lycopene degradation needed to be addressed. Nine of
the replicate purees and nine of the replicate chunk samples for each

10.1021/jf030022f This article not subject to U.S. Copyright. Published 2003 by the American Chemical Society
Published on Web 05/01/2003
Watermelon Lycopene Stability J. Agric. Food Chem., Vol. 51, No. 12, 2003 3583

individual watermelon were frozen at -20 °C. Similarly, nine puree Table 1. The Effect of Freeze−Thaw Cycles on Lycopene Content of
replicates and nine chunk replicates for each watermelon were frozen Watermelon Tissue Samples
at -80 °C. The remaining two replicates of fresh tissue from each
watermelon were assayed for lycopene, and their average served as number of lycopene content (% of fresh tissue)a
the lycopene content of fresh tissue from each fruit and against which freeze−thaw puree puree chunks chunks
the lycopene contents of the frozen samples were compared. The cycles −20 °C −80 °C −20 °C −80 °C
lycopene contents of fresh heart tissue from the nine individual
watermelons ranged from 42.9 mg/kg to 78.2 mg/kg. The average one 96.7 ± 1.2 97.4 ± 1.2 93.6 ± 3.3 92.9 ± 4.3
two 96.0 ± 1.0 96.1 ± 1.4 92.1 ± 4.8 92.3 ± 4.6
lycopene content of the nine watermelons was 53.6 ( 10.8 mg/kg (mean
( s.d.), and the median lycopene content was 53.2 mg/kg. Samples
were stored for various lengths of time at -20 °C or -80 °C. For the
a Average ± standard deviation for samples from nine watermelons.
long-term storage study, each sample was thawed only once before
analysis. Thawed samples were pureed using a Brinkmann Polytron
Homogenizer (Brinkmann Instruments, Inc., Westbury, New York) with
a 20-mm O. D. blade to produce a uniform slurry with particles smaller
than 3 mm3. The samples were not allowed to heat or froth during
homogenization by chilling them on ice and using short durations of
homogenization.
Low-Volume Hexane Extraction Quantitative Analysis of Lyco-
pene. The low-volume hexane extraction method was performed as
described in Fish et al. (11). Once pureed, samples were stirred on a
magnetic stirrer during sampling for analysis. Approximately 0.6-g
(determined to the nearest 0.01 g) duplicate samples were weighed from
each puree into two 40-mL amber screw-top vials (Fisher, #03-391-
8F) that contained 5 mL of 0.05% (w/v) butylated hydroxytoluene
(BHT) in acetone, 5 mL of 95% USP grade ethanol, and 10.0 mL of
hexane. Samples were extracted on an orbital shaker at 180 rpm for 15
min on ice. After shaking, 3 mL of deionized water was added to each
vial, and the samples were shaken for an additional 5 min on ice. The
vials were then left at room temperature for 5 min, to allow for phase
separation. The absorbance at 503 nm of the upper hexane layer was
measured in a 1-cm path length quartz cuvette after zeroing the Figure 1. The rate of loss of lycopene during frozen storage of watermelon
instrument with pure hexane. The lycopene content of each sample tissue. Lycopene levels for nine samples were normalized to percent of
was then estimated using the absorbance at 503 nm and the sample the fresh level, and the average of the nine samples for each time point
weight (11, 12). was plotted versus the time of storage. Open circles, puree stored at
Fitting of Experimental Data With Kinetic Equations. Our data −20 °C; solid circles, puree stored at −80 °C; open triangles, chunks
for lycopene deterioration during frozen storage could not be fitted stored at −20 °C; solid triangles, chunks stored at −80 °C. Best fit lines
with simple, single-step rate equations. We were able, however, to fit were generated for the corresponding set of data points by eqs 1, 2, or
the data by resolving them into two first order decays of the general
3 in the text.
form
Ct/Co ) P1 exp(-k1t) + P2 exp(-k2t) evaluated by thawing a chunk sample and a puree sample from
each watermelon after the samples had been frozen for 2 days
In this equation, Ct is the concentration of lycopene in tissue after and assaying for lycopene content. The unused portion of each
storage for time, t, in days. Co is the concentration of lycopene in fresh
sample was then refrozen and assayed again about 24 h later to
tissue minus 4% for freeze-thaw destruction. The rate constants, k1
and k2, are the respective first-order rate constants for each of the two
evaluate the effect of a second freeze-thaw cycle. The lycopene
first-order processes. P1 and P2 are constants for each process and are content of flesh from each melon expressed as a percent of the
related to lycopene measurement and the proportions of lycopene lycopene content of the fresh tissue was averaged over the nine
degrading by each process. The value of the smaller rate constant, k2, melon samples (Table 1). Inspection of the standard deviations
(slower rate) was determined from the slope of the linear portion of a of the treatments reveals that the puree form produces a more
plot of ln(Ct/Co) versus t at large values of t. The intercept upon uniform set of samples than does the chunk form (s.d. for puree
extrapolation to t ) 0 yielded values for P1 and P2 (because P1 + P2 ∼1.2; s.d. for chunks ∼4.3). Analysis of variance indicated that
) 1) (13). The contribution of P2 exp(-k2t) (slower process) to Ct/Co there was a significant difference (p < 0.05) between the frozen-
was then subtracted from the left-hand side of its corresponding thawed samples and the fresh tissue. However, there was no
equation, and the ln of the residual was plotted against time. These
significant difference (p > 0.05) in the means of once frozen-
plots were linear, within the uncertainty of the data, and their slopes
yielded the value of k1, the larger rate constant (faster rate).
thawed and twice frozen-thawed or between puree and chunk
HPLC Analysis of Carotenoids. The distribution and relative levels samples. We were unable to determine the cause of this ∼4-
of trans-lycopene, its collective geometric isomers, and other carotenoids 6% decrease in lycopene as a result of freeze-thaw. We have
in several selected samples were determined by HPLC and diode array observed this behavior previously (15) but did not report the
detection by the method of Tonucci et al. (14). This method does not amount of degradation.
resolve the individual cis-isomers, but collectively separates them from Lycopene Levels with Temperature and Time of Frozen
all-trans-lycopene. Storage. Samples from each watermelon were removed from
Statistical Analysis. Statistical analyses that included analysis of frozen storage at selected intervals over a year’s time, and each
variance, mean, and standard deviation determinations were performed was assayed for lycopene content. The results are summarized
with the aid of Statistica software, version 6 (StatSoft, Tulsa, OK)
in Figure 1. Three observations are apparent from the results.
First, lycopene in frozen watermelon tissue is more stable when
RESULTS
stored at -80 °C than when stored at -20 °C. Second, again it
Effect of Freeze-Thaw on Lycopene Levels. The potential is suggested by comparison of the scatter in the respective data
effect of a freeze-thaw process on watermelon tissue was that the puree form of tissue produces the more homogeneous
3584 J. Agric. Food Chem., Vol. 51, No. 12, 2003
set of samples. Part of the source of variation in the chunk
samples may have occurred by loss of variable amounts of water
out of the chunks and into the plastic bags. Third, the form in
which the tissue was stored has little effect on lycopene stability
compared to the effect of temperature, because chunks and puree
behave basically identically at -80 °C. However, at -20 °C,
an apparently faster rate of lycopene deterioration was exhibited
by chunks than by puree.
We attempted to fit the lycopene deterioration data with
simple, single-step rate equations. The results could not be
adequately fit with the linearized forms of zero-, first-, or
second-order rate equations; in all cases, curvilinear plots were
obtained. This was true whether the data were adjusted for the
∼4% lycopene loss by freeze-thaw or not. The results could,
however, be resolved into two first-order decay processes. The
equations are given below for three of the cases.
Chunks, -20 °C: Ct/Co )
0.15 exp(-0.03t) + 0.85 exp(-0.0008t) (1)
Puree, -20 °C: Ct/Co )
0.15 exp(-0.018t) + 0.85 exp(-0.00045t) (2)
Puree, -80 °C: Ct/Co )
0.04 exp(-0.04t) + 0.96 exp(-0.000045t) (3)
Each line in Figure 1 was generated by its corresponding
equation and offers a reasonable fit to the experimental data.
No attempt was made to fit the data for chunks at -80 °C,
because it was similar to that for the puree but was less precise.
The fact that the experimental data may be satisfactorily
described by a combination of two first order processes does
not prove that this is actually what is occurring during storage
and certainly does not imply a mechanism of degradation. The
data do imply, however, that more than one deterioration process
is likely operating on lycopene during its frozen storage.
Equations 1-3, rather, provide a means to predict the amount
of lycopene degradation expected for storage of frozen water-
melon tissue at one of two temperatures over a given period of
time.
Additionally, we evaluated if any portion of all-trans-lycopene
was being converted to its collective forms of cis-isomers during
storage. HPLC separation of the carotenoids from -20 °C and
-80 °C puree samples from four watermelons was performed
after one year of storage. No measurable increase in the
collective quantities of cis-lycopene isomers was observed. One
interesting observation, however, was that samples stored at -20
°C exhibited a ratio of β-carotene to lycopene about 10 percent
greater than that of its counterpart stored at -80 °C. This held
true for the four samples whose two storage temperatures were
compared in this manner. The results suggest that deterioration
of β-carotene is slower at -20 °C than is that of lycopene. This
differential rate of degradation of lycopene and β-carotene
should have no effect on the quanitative spectral assay for
lycopene, because of the widely disparate levels of the two in
watermelon, ca. 15 times more lycopene than β-carotene (11).
DISCUSSION
Several factors need to be considered if reliable estimates of
lycopene content of fresh watermelon tissue are to be derived
from tissue that has been frozen and stored for a period of time
before assay. First, a more homogeneous sample will be obtained
if the tissue is homogenized before sampling and storing. This
will result in fewer replicates needed to achieve a requisite
precision level in the assay results. Second, the pureed form is
Fish and Davis
at least as stable as the chunk form, and the lycopene appears
to degrade more slowly at -20 °C in the puree than in the chunk
form. Third, we observed a loss of ∼4-6% of the watermelon
tissue lycopene as a result of a single freeze-thaw cycle. We
hypothesize that this more labile form of lycopene may represent
a small pool of tissue lycopene that is either environmentally
or physiologically in an unstable state and is oxidized or
otherwise degraded upon freezing-thawing. This small, but
consistent, loss of lycopene has not been observed, to our
knowledge, in the tomato system. It was also reported that both
lycopene and β-carotene in liquid-frozen human serum remained
stable through repeated freeze-thaw cycles (16). Together, these
results suggest that this labile pool of lycopene does not exist
in some systems or that the tissue matrix in watermelon does
not afford the same level of protection as do the matrixes in
tomato. For watermelon tissue, it appears that one should correct
for this fractional loss of the fresh tissue lycopene when dealing
with a frozen-thawed sample. Fourth, the temperature of storage
exerts a substantial effect on the stability of lycopene in frozen
watermelon tissue. Not surprisingly, the lower the temperature,
the more stable the lycopene. Although -80 °C is the preferred
storage temperature, realistically, such cold storage conditions
will not be commonly available, and -20 °C will have to
suffice. Figure 1 or calculation with eq 2 shows that about 10
% (in addition to the ∼4% freeze-thaw loss) of the lycopene
will degrade after 30 days storage of watermelon tissue puree
at -20 °C, while less than this is lost after a year at -80 °C.
Thus, storage time at -20 °C should be as short as possible, to
minimize the uncertainty of the analysis. No shift was observed
in the amount of all-trans-lycopene isomer to the collective cis-
isomers during frozen storage. This is not totally unexpected
in light of ab initio molecular modeling computational studies
that indicate that all-trans-lycopene is more stable than its cis
isomers, with the exception of 5-cis lycopene (17).
Because this storage study was conducted in the presence of
an oxygen atmosphere (no effort was made to remove O2 from
the puree or from the atmosphere above the chunks or puree),
it is highly likely that the degradation of lycopene occurred via
oxidative processes. It is well documented that caratenoids can
be oxidatively cleaved to form aldehydes, ketones, epoxides,
and endoperoxides (18-21). Frozen storage of watermelon
tissue in the absence of oxygen may slow the rate of lycopene
degradation, but will likely not entirely eliminate the degrada-
tion. For example, tomato pulp samples stored in the dark at
-20 °C still lost lycopene at one-half the rate in vacuuo as in
the presence of air (5). Thus, ∼16% of the tomato pulp lycopene
was lost during 60 days storage at -20 °C in the absence of
O2. In a model aqueous system, about 60% of the lycopene
deteriorated after 7 h at 30 °C under a N2 atmosphere (18).
Our data for frozen watermelon tissue are consistent with at
least two lycopene degradation processes that occur simulta-
neously. The results may be described by two simultaneous first
order decays (eqs 1-3), where one process is about forty times
faster than the other. Obviously, this is a phenomenological
kinetic description of what we observe and offers no mechanistic
insight into what is actually happening. The rates and mecha-
nisms of the degradation of lycopene in the presence of oxygen
appear to be system dependent. Sharma and LeMaguer (5)
observed that the degradation of lycopene in tomato pulp at
-20, 5, and 25 °C could be described by a single, pseudo-first-
order reaction at each of the three temperatures. Henry et al.
(18) observed that the degradation of lycopene exhibited zero-
order kinetics when adsorbed on a C18 solid phase in an aqueous
system in the presence of O2 or ozone at room temperature.
Watermelon Lycopene Stability
Agarwal et al. (8) observed no significant loss of lycopene in
tomato juice stored at 4 °C for 12 months. Together, these results
support the contention that the observed behavior of lycopene
degradation will depend heavily on the medium in which it is
stored and emphasizes that our results for watermelon tissue
should not be extrapolated to lycopene stored in other water-
melon systems (eg., juice).
Our observation that, in watermelon tissue, lycopene appears
to degrade more rapidly than β-carotene during frozen storage
is consistent with the observations of Lee and Coates (22) in a
different food system. They observed >20% loss of lycopene
and a 7% loss in β-carotene in pink grapefruit juice stored at
-23 °C in plastic containers for 12 months. These observations
with food systems are consistent with observations of in vitro
model systems, in which the decreasing number of coplanar
conjugated double bonds and the presence of hydroxy and keto
groups in carotenoids decrease their reactivity in radical
scavaging reactions (23, 24).
ACKNOWLEDGMENT
We would like to thank Rick Houser, and Anthony Dillard for
providing valuable technical support. We gratefully acknowledge
the HPLC analyses provided for us by Julie Collins. We also
acknowledge the provision of the watermelons by Warren
Roberts and Wyatt O’Hern of Oklahoma State University.
LITERATURE CITED
(1) Sandmann, G. Carotenoid biosynthesis in microorganisms and
plants. Eur. J. Biochem. 1994, 223, 7-24.
(2) Di Mascio, P.; Kaiser, S. P.; Sies, H. Lycopene as the most
efficient biological carotenoid singlet oxygen quencher. Arch.
Biochem. Biophys. 1989, 274, 532 538.
(3) Gerster, H. The potential role of lycopene for human health. J.
Am. Coll. Nutr. 1997, 16, 109-126.
(4) Arab, L.; Steck-Scott, S.; Bowen, P. Participation of lycopene
and Beta-carotene in carcinogenesis: defenders, aggressors, or
passive bystanders? Epidemiol. ReV. 2001, 23, 211-230.
(5) Sharma, S. K.; Le Maguer, M. Kinetics of lycopene degradation
in tomato pulp solids under different processing and storage
conditions. Food Res. Int. 1996, 29, 309-315.
(6) Shi, J.; Le Maguer, M. Lycopene in tomatoes: chemical and
physical properties affected by food processing. Crit. ReV.
Biotechnol. 2000, 20, 293-334.
(7) Thompson, K. A.; Marshall, M. R.; Sims, C. A.; Wei, C. I.;
Sargent, S. A.; Scott, J. W. Cultivar, maturity, and heat Treatment
on lycopene content in tomatoes. J. Food Sci. 2000, 65, 791-
795.
(8) Agarwal, A.; Shen, H.; Agarwal, S.; Rao, A. V. Lycopene content
of tomato products: its stability, bioavailability and in vivo
antioxidant properties. J. Med. Food 2001, 4, 9-15.
(9) Collins, J. K.; Perkins-Veazie; P. M. Lycopene stability in stored
and cut watermelon. Institute of Food Technologists Annual
Meeting, New Orleans, LA, 2001, Program Abstract 44C-16, p
90.
J. Agric. Food Chem., Vol. 51, No. 12, 2003 3585
(10) Motes, J.; Cuperus, G. Cucurbit production and pest management.
OK Coop. Ext. Serv. Circular E-853, 1995.
(11) Fish, W. W.; Perkins-Veazie, P.; Collins, J. K. A quantitative
assay for lycopene that utilizes reduced volumes of organic
solvents. J. Food Compos. Anal. 2002, 15, 309-317.
(12) Beerh, O. P.; Siddappa, G. S. A rapid spectrophotometric method
for the detection and estimation of adulterants in tomato ketchup.
Food Technol. 1959, 13, 414-418.
(13) Ikai, A.; Tanford, C. Kinetics of unfolding and refolding of
proteins. I. Mathematical analysis. J. Mol. Biol. 1973, 73, 145-
164.
(14) Tonucci, L. H.; Holden, J. M.; Beecher, G. R.; Khachik, F.,
Davis, C. S.; Mulokon, G. Carotenoid content in thermally
processed tomato-based food products. J. Agric. Food Chem.
1995, 43, 579-586.
(15) Davis, A. R.; Fish, W. W.; Perkins-Veazie, P. A rapid hexane-
free method for analyzing lycopene content in watermelon. J.
Food Sci. 2003, 68, 328-332.
(16) Thomas, J. B.; Duewer, D. L.; Kline, M. C.; Sharpless, K. E.
The stability of retinol, R-tocopherol, trans-lycopene, and trans-
β-carotene in liquid-frozen and lyophilized serum. Clin. Chem.
Acta 1998, 276, 75-87.
(17) Chasse, G. A.; Mak, M. L.; Deretey, E.; Farkas, I.; Torday, L.
L.; Papp, J. G.; Sarma, D. S. R.; Agarwal, A.; Chakravarthi, S.;
Agarwal, S.; Rao, A. V. An ab initio computational study on
selected lycopene isomers. J. Mol. Structure THEOCHEM 2001,
57, 27-37.
(18) Henry, L. K.; Puspitasari-Nienaber, N. L.; Jaren-Galan, M.; van
Breemen, R. B.; Catignani, G. L.; Schwartz, S. J. Effects of ozone
and oxygen on the degradation of carotenoids in an aqueous
model system. J. Agric. Food Chem. 2000, 48, 5008-5013.
(19) Handelman, G. J.; van Kuijk, F. J.; Chatterjee, A.; Krinsky, N.
I. Characterization of products formed during the autoxidation
of β-carotene. Free Radical Biol. Med. 1991, 10, 427-437.
(20) Marty, C.; Berset, C. Factors affecting the thermal degradation
of all-trans-β-carotene. J. Agric. Food Chem. 1990, 38, 1063-
1067.
(21) Stratton, S. P.; Schaefer, W. H.; Liebler, D. C. Isolation and
identification of singlet oxygen oxidation products of β-carotene.
Chem. Res. Toxicol. 1993, 6, 542-547.
(22) Lee, H. S.; Coates, G. A. Characterization of color fade during
frozen storage of red grapefruit juice concentrates. J. Agric. Food
Chem. 2002, 50, 3988-3991.
(23) Woodall, A. A.; Lee, S. W.-M.; Weesie, R. J.; Britton, G.
Oxidation of carotenoids by free radicals: relationship between
structure and reactivity. Biochim. Biophys. Acta 1997, 1336, 33-
42.
(24) Mortensen, A.; Skibsted, L. H. Importance of carotenoid structure
in radical-scavenging reactions. J. Agric. Food Chem. 1997, 45,
2970-2977.
Received for review January 10, 2003. Revised manuscript received
March 17, 2003. Accepted for publication March 19, 2003. Mention of
trade names or commercial products in this article is solely for the
purpose of providing specific information and does not imply recom-
mendation or endorsement by the U.S. Department of Agriculture.
JF030022F